By Agnieszka Gautier
Some of the greatest scientific discoveries have happened by chance or when things do not go to plan. One such example of scientists turning potential disaster into opportunity occurred after the launch of NASA’s Soil Moisture Active Passive (SMAP) satellite in early 2015.
Just two months after SMAP launched, the satellite’s radar transmitter failed, rendering it unable to send out a wave signal. The frequency of the transmitter, however, was similar to the Global Positioning System (GPS) satellite signals, so it could still potentially receive data. From Earth, the mission team moved the receiver’s frequency to the GPS L2 band, one of its three frequencies. It worked. Starting in 2015, data from SMAP were collected using GPS signals. They had been stored, but not processed, because to do so required expertise in Global Navigation Satellite System-Reflectometry (GNSS-R), a specific niche in the science community.
Fortunately, NASA’s open science and data policy allows for collaboration between diverse scientific groups, which paved the way for Nereida Rodriguez-Alvarez, a specialist on GNSS-R techniques, to eventually jump in and help make sense of the collected data.
“For years, no one knew what Earth surface information SMAP held,” said Rodriguez-Alvarez, a scientist at NASA’s Jet Propulsion Laboratory. After downloading the raw SMAP data, which is stored at the NASA National Snow and Ice Data Center (NSIDC) Distributed Active Archive Center (DAAC), Rodriguez-Alvarez and her team needed to get the signal out from the raw data. “Once we began processing, we realized we had very unique measurements, unlike anything we had ever seen,” she said, referring to a study her team had published in March 2023. The SMAP L1B Polarimetric GNSS Reflectometry data set, described within the study, is now available at the NSIDC DAAC and makes the full polarimetric calibrated data accessible.
Crossing new paths
To grasp the challenges behind processing the data, it is important to understand how the satellite was meant to function in the first place. The SMAP satellite houses two instruments, an active radar and a passive radiometer, which work together to measure soil moisture and distinguish between frozen and thawed ground. The radiometer measures natural radiation emitted from Earth’s surface, while the radar sends a signal and waits for its bounce back.
Throughout its lifespan, the radiometer on the satellite always worked as intended. However, when the active radar’s transmitter suffered a glitch in 2015, it was no longer able to send out a signal, but still able to receive information. Normally, SMAP would transmit and receive a signal in a straight line, but GPS signals move in a circular pattern. So SMAP captured the data in an unexpected way.
By having SMAP’s active radar receive GPS signals, scientists had created a bistatic radar, where the transmitter and receiver are located away from one another. Multiple advantages came from this setup, including how the receiver received the signal.
Less bounce, more accuracy
In analyzing radar signals, backscatter and forward scatter are two terms used to describe how electromagnetic waves interact with objects in their path. Backscatter occurs when the signal bounces back to the source, while forward scatter occurs when the signal disperses away from the source.
Backscatter in radar often suffers from double-bounce reflections, which reflect off of two or more surfaces before returning to the receiver. Imagine a double-bounce radar signal like a ball passing through the atmosphere, hitting the ground, and then bouncing off a tree before ricocheting back out through the atmosphere to the receiver. Because the signal bounces off unintended objects, this double bounce complicates scientists’ ability to extract accurate soil moisture measurements.
However, forward scatter has a lower-double bounce signal because of its more advantageous geometry. Because of the angle at which a signal is sent, it is less likely for that signal to bounce off a tree and down to Earth’s surface and back away. In densely vegetated areas, forward scatter can therefore minimize double-bounce effects. “So, what you end up getting is more accurate soil moisture data in those regions,” said Rodriguez-Alvarez, who published the findings in Nature.
This unique configuration also allowed SMAP to measure the circularly reflected GPS signals in both horizontal and vertical orientations, known as polarizations. Now SMAP has two views, or two 90-degree angles, for every data point, rather than one.
When you have the two signals, like the horizontal and vertical, you can better distinguish soil and dense vegetation. But how exactly did Rodriguez-Alvarez extract the information from the data?
The process of processing the data
Processing the data was not easy. It required a lot of research, math, and creativity.
“Not only was it not easy to recover the GPS signal from the raw data, but it was also not easy understanding what we were measuring,” said Rodriguez-Alvarez. “We come from another way of doing things.”
As scientists who are part of the global navigation community, the team was used to the GPS circular polarization. The complication with the GPS-SMAP relationship is that the transmitter sent out a circular signal while the receiver measured the return signal at two right-angle polarizations. The team had to do a lot of research on these hybrid signals. Luckily, Synthetic Aperture Radar (SAR), commonly used to scan Earth’s surface, works with a similar hybrid model but in backscatter mode. “So, we took the SAR equations and adapted them to the forward scatter,” she added.
Once the team realized that SMAP radar enabled two additional independent sources of information, rather than one, they needed to figure out where the signals were coming from, especially in those densely vegetated areas where radar gets bouncy. They used a model that can accommodate the active and passive components of SMAP. The team could ingest four measurements, two from SMAP and two from GPS, into the model to retrieve vegetation, soil moisture, and soil roughness to get better results.
The team has processed data from 2015, and as of September 2024, the data are updated monthly and available from the NSIDC DAAC. Data are fully compatible with other global navigation missions like the format used for the Cyclone Global Navigation Satellite System (CYGNSS) as most GNSS-R missions are now converging on a single standard. “The GNSS-R community is very used to that format. Our safest option is to follow that format so that the GNSS-R community can jump in and explain that format very well to others,” said Rodriguez-Alvarez.
An unexpected discovery
While processing the data, an unexpected discovery unveiled itself. While mapping out the radar pathways, Rodriguez-Alvarez and her team discovered two clear signatures over Greenland.
“This is one area on the planet that I do not have a lot of experience working with,” she said. “The signatures over Greenland surprised me.” One signature is related to snowmelt, while the other measures ice sheet thickness.
Figuring out what measurements have been stored within SMAP’s collected data was the first step in unlocking its potential.
“At the beginning of the project, I was not expecting to do something so complex,” Rodriguez-Alvarez said. “But we built this mathematical formulation to get the measurements, and now it all makes sense.”
Now scientists are able to use the data for purposes far beyond soil moisture and freeze thaw research. As Rodriguez-Alvarez and her team have proven, there is still a lot to unlock from this serendipitous meeting of GPS and SMAP.
About the data
Data set: SMAP L1B Polarimetric GNSS Reflectometry data set
This product represents the first full polarimetric Global Navigation Satellite System Reflectometry (GNSS-R) dataset, derived from the Soil Moisture Active Passive (SMAP) mission radar. The main parameters are the normalized Stokes parameters, total power normalized bistatic radar cross section, and reflectivity.
The NSIDC DAAC also archives a variety of data sets on behalf of NASA, for SMAP and other data collections. In addition to archiving data, the NSIDC DAAC provides data users with multiple methods of support: documentation, help articles, data tools, training, and on-demand support.
References
Munoz-Martin, J. F., Rodriguez-Alvarez, N., Bosch-Lluis X. and K. Oudrhiri. 2022. Stokes Parameters Retrieval and Calibration of Hybrid Compact Polarimetric GNSS-R Signals. IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-11, 2022, Art no. 5113911, doi:10.1109/TGRS.2022.3178578.
Rodriguez-Alvarez, N., J. F. Munoz-Martin, X. Bosch-Lluis, and K. Oudrhiri. 2024. SMAP L1B Polarimetric GNSS Reflectometry. (NSIDC-0795, Version 1). [Data Set]. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center, doi:10.5067/ZF4OXSD2NS7W.
Rodriguez-Alvarez, N., Munoz-Martin, J. F., Bosch-Lluis X. and K. Oudrhiri. 2023. A Hybrid Compact Polarimetry GNSS-R Analysis of the Earth’s Cryosphere. IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1-15, 2023, Art no. 4301615, doi:10.1109/TGRS.2023.3280363.
Rodriguez-Alvarez, N., Munoz-Martin, J. F., Bosch-Lluis, X. et al. 2023. The first polarimetric GNSS-Reflectometer instrument in space improves the SMAP mission’s sensitivity over densely vegetated areas. Nature Scientific Reports 13, 3722. doi:10.1038/s41598-023-30805-7.